Transparent Conductors: Part I

Prof. Tom Mason Northwestern University

Materials Science & Engineering

• Some History and

Applications • Basic requirements for TCOs • Role of Crystal Structure • Phase Diagrams • Bulk Experimental Methods:

• Synthesis • Characterization

• Defect Chemistry

~

Transparency (> 80% through visible spectrum for

typical 1 µm thin film)

Conductivity (> 103 S/cm)

An Introduction to TCOs

TCO History and Applications

“a brown/black powder” σ = 37 S/cm

F. Streintz, Ann. Phys. (Leipzig) 9, 854 (1902)

The First Transparent TCO

CdO

thermal oxidation of sputtered Cd films on glass σ =870 S/cm (106 nm thick) with orange/gold color

K. Bädeker, Ann. Phys. (Liepzig) 22, 749 (1907)

The First Thin Film TCO

Mid-20th Century Developments

• 1930s-1940s: conductive SnO2 film patents (various glass companies)

• 1951: first ITO patent (Corning) Sn-doped In2O3

• 1959: key dissertation on ZnO properties

• 1971: ZnO varistor and TCO film patents (Japan) SnO2-based windshield de-icer

Early TCO Science: In-doped ZnO

Single crystal data: E. Scharowsky, A. Physik, 135, 318 (1953) R. Arneth, PhD thesis, U. Erlangen (1959) (taken from: G. Heiland, E. Mollwo and F. Stockmann, in Solid State Physics, Vol. 8, ed. By F. Seitz and D. Turnbull, 1959)

TCO Development (1970-2000)

Source: T. Minami, MRS Bulletin, August, 2000 -data for pure and doped host oxides.

Why the Surge of Interest?

• Novel complex oxide/solid solution TCOs discovered in the 1990s

• Discovery of p-type TCO materials since the late 1990s

• Development of amorphous n-type TCOs since 2004

• Large area applications (organic LEDs, solar cells) require ITO-alternatives (chemical, electrical, cost issues)

• Development of Transparent Oxide Semiconductors (oxide-based thin film transistors-TTFTs)

• Low-emissivity windows • Transparent front

electrodes for flat-panel displays

• Transparent top electrodes for photovoltaic cells

• Defrosting windows (freezers, cockpits)

• Electrochromic mirrors and windows

• Oven window coatings • Static charge dissipation

coatings • Touch-panel controls • Electromagnetic shielding • Invisible security circuits • Organic light-emitting

diodes - R. Gordon, MRS Bulletin, August 2000.

A Wide Array of TCO Applications

Large-Area Applications • High electrical conductivity (>1000 S/cm) • High optical transparency in visible region (>80%) • Industry standards: ITO, SnO2

Ga2O3

In2O3

ZnO

CdO

SnO2

Basic TCO Requirements

Basic TCO Requirements

I. Hamberg and C. G. Grandqvist, J. Appl. Phys., 60, R123 (1986).

Eg0 > 3.1 eV

N-type Degenerate Doping

Eg > Eg0

Insulating Parent Compound

• Parent oxide with relatively wide band gap • Interband transitions > 3.1 eV (cations with filled d-shells) • Highly dispersed conduction (or valence) band (high mobility) • Ability to donor- or acceptor-dope to ~1021/cm3

TCO-Active Species: d10 Cations

Cu Zn Ga

Ag Cd In Sn

= p-type = n-type

+1 +2 +3 +4 valence state

Typical Parameters:

• Large electron populations in the 1020 to 1021 cm-3 range (highly degenerate)

• High electron mobilites in the 30-70 cm2/V-s range

• Large conductivities (in excess of 1000 S/cm) • Low absorption in the visible (α<104 cm-1)

– 100nm film 90% transparent in the visible – 500nm (0.5µm) film 60% transparent in the visible

Role of Crystal Structure

Shannon’s n-type TCO Maxim

“…continuous edge-sharing of Cd2+, In3+ and Sn4+ octahedra is a necessary criterion for the formation of an n-type transparent conductor.”

-R. D. Shannon et al., J. Phys. Chem. Solids, 38, 877 (1977).

Octahedral Cation TCO Family

CdO SnO2

rocksalt structure rutile structure

Bixbyite: a Fluorite-Derivative

• Double ZrO2 to Zr2O4

• If In-oxide were a fluorite, it would be In2O4

• But In3+ rather than In4+

• For charge balance, remove ¼ of O-ions

• Get In2O3

Octahedral Cation TCO Family

• Edge-shared octahedra • The octahedra are

distorted (hints at why amorphous TCOs also conduct)

• Can be readily doped (e.g., Sn4+) and co-doped (Zn2+/Sn4+ and Cd2+/Sn4+ )

• The basis of a rich family of complex oxide solid solutions

In2O3

bixbyite structure

Octahedral Cation TCO Family

Yellow: Tetrahedral (Td)

Blue: Octahedral (Oh)

spinel structure

• Spinel = MgAl2O4 • 1/2 octahedral

interstices occupied: octahedral sites

• 1/8 of tetrahedral interstices occupied: tetrahedral sites

• Arrays of edge-shared cation octahedra

• Spinels are well-known for exhibiting extended solid solutions

The Importance of Phase Diagrams

The Cd-In-Sn-O (CITO) System

1175°C

CdO (rocksalt) o-Cd2SnO4

(Sr2PbO4) CdSnO3

(o-Perovskite)

CdIn2O4 In2-2x(Cd,Sn)2xO3 (0 < x < 0.34)

(bixbyite)

ITO

SnO2 (rutile)

(1-x)CdIn2O4 – (x)Cd2SnO4 (0 < x < 0.75)

(spinel)

InO1.5 (bixbyite)

rutile + bixbyite

spinel + rocksalt

spinel + bixbyite

2+

3+

4+

Optimized conductivity vs. OSD

0

5

10

15

20

0 1 2 3 4 5

o-C

dSnO

3

o-C

d2Sn

O4

Cd2

SnO

4

Sn-d

oped

In2O

3

In-d

oped

CdO

σ (k

S/cm

)

oct-site density (x1022 cm-3)

SnO

2

Bellingham et al. (1991) predicted the maximum TCO conductivity = 25,000 S/cm

Shannon’s n-type TCO Maxim

“…continuous edge- or corner- sharing of Cd2+, In3+ and Sn4+ octahedra is a necessary criterion for the formation of an n-type transparent conductor.”

-R. D. Shannon et al., J. Phys. Chem. Solids, 38, 877 (1977).

What about Zn2+?

In2O3(ZnO)k

T = 1275 ˚C

Zn2SnO4

The Zn-In-Sn-O System

spinel conductivities lo

g(σ)

Mole fraction CdIn2O4 ZnIn2O4

Cd2SnO4 Zn2SnO4

-2

-1

0

1

2

3

4

5

0.0 0.2 0.4 0.6 0.8 1.0

G.P. Palmer, K.R. Poeppelmeier, and T.O. Mason. J Solid State Chem 134. 192 (1997) D.R. Kammler, T.O. Mason, D.L. Young, and T.J. Coutts. J App Phys 90. 3263 (2001) X. Wu, T.J. Coutts, and W.P. Mulligan, J Vac Sci Tech A15. 1057 (1997) D.R. Kammler, T.O. Mason, and K.R. Poeppelmeier. Chem Mater 12. 1954. (2000)

(1-x)ZnIn2O4•xZn2SnO4

(1-x)CdIn2O4•xCd2SnO4

Bulk

Bulk

Thin film

x = 1 Not effective: Zn2+ (oct) Zn2+ (tet)

What About Zn2+ ?

Shannon’s n-type TCO Maxim

“…continuous edge- or corner- sharing of Cd2+, In3+ and Sn4+ octahedra is a necessary criterion for the formation of an n-type transparent conductor.”

-R. D. Shannon et al., J. Phys. Chem. Solids, 38, 877 (1977).

What about ZnO?

Tetrahedral Coordination TCO

• ZnO is the only TCO with exclusively tetrahedral coordination

• In other crystal structures/CNs Zn is not as strong a contributor to TCO behavior

• Zn cations play an important role in stabilizing amorphous n-type TCOs

wurtzite structure

ZnO

Bulk Experimental Methods

• Batch calculations of precursors (usually constituent oxides) • Weighing and mixing of precursors

1. Obtain starting oxide

powders

2. Dry powders

3. Batch Calculations

4. Weigh out correct

quantities

5. Grind in acetone

Solid State Reactions

Cold pressing Hot pressing

• Al2O3 dies prevent reduction • 67 – 774 MPa max • 0.5” to 1.7” pellets

• Typically ~130 MPa with steel dies • Can add binder/ pressing aid • 0.25” to 1.5” pellets

Pellet Pressing

• Quenching preserves high temperature phase Limitation: thermal shock

• Nested crucibles minimize preferential evaporation (e.g. Zn) • Sacrificial powder minimizes contamination

• Up to ~ 1400°C • In fume hoods when needed

Time frame: 1 day – 2 weeks

Solid State Reaction/Sintering

Computer Control

Lens

Sapphire Window

Heated Substrate

Holder

Excimer Laser

λ=248nm

Vacuum Pumps

Rotated To Prevent

Localized Heating

Stepper Index

Between Targets

Reaction Gas Inlet

Multi-target Pulsed Laser Deposition (collaboration with Prof. Chang, Dr. Buchholz)

Rietveld Analysis

35

Weight Fraction of Each Phase:

Determining Phase Boundaries

36

Lever Rule: fβ = (y-z)/(x-z) Disappearing Phase Method

Phase Boundaries: Vegard's Law

37

• Lattice constant changes linearly with composition until the solubility limit is reached. • Can use Rietveld refinement with internal silicon standard to accurately measure the lattice constant.

898°C

894°C

R

R+S

S S+W

R+W

C=0.82

C=Co/(Zn+Co)

C=0.77

T

The Zinc Oxide – Cobalt Oxide System

T-X Phase Diagram in Air

4-Point Conductivity and Thermopower

39

In situ bulk electrical measurements

Conductivity vs. time for change of pO2 (from 1000ppm to 1%O2) at 750oC

Controlled atmosphere electrical measurements performed for a sample of c-ZITO (In1.2Sn0.40Zn0.40O3)

PC-controlled current source and multimeter

Gold electrodes

Sample

Alumina sample and TC holder

log pO2

controlled atmosphere tube furnace:

log pO2

p-type n-type

Slopes majority defect type

Brouwer Analysis for Determining Point Defect Mechanism

Powder-Solution-Composite Method Slurry: NaCl solution + powder sample

Stainless steel electrodes Polypropylene tube

Comparison of slurry composite conductivity to matrix (plain NaCl solution) conductivity: Cross over point corresponds to bulk value for powder in question. Use effective medium theory to fit the data.

Motivation: Measurement of low-temperature derived materials in powder form. Technique: Based on impedance spectroscopy of composite materials

AC Impedance Spectroscopy (AC-IS)

43

Apply oscillating voltage, measure current

response.

Ratio is the complex impedance, which depends on

frequency.

Nyquist plot can illustrate different processes in

frequency domain

Phase offset = θ/ω

θ

Hall Effect

44

RH = Hall coefficient Hall Mobility:

Hall Effect: 17 T and 0.58 T

45

17 T

0.58 T

Both Hall resistance and Seebeck coefficient are positive

pedBRxy +=

+

+= A

pNLn

ekQ v

46

+Ni

+Mg

Co2+xZn1-xO4

Experiment: Confirms Robust P-type Character

Diffuse Reflectance Spectroscopy

47

Sample

Detector

Integrating sphere

Diffuse Reflectance Spectroscopy

Zn-rich Rh2ZnO4

Defect Chemistry of n-type TCOs

Kröger-Vink Notation:

Ms c

Matter (or lack thereof)

Site (or whole crystal, if electrons)

Effective charge (the charge relative to the perfect crystal) x = neutral • = positive ' = negative

Writing Balanced Defect Reactions:

Ms c

Mass Balance (massleft = massright)

Site Ratio (host site ratioleft = host site ratioright)

Charge Balance (chargeleft = chargeright)

Note: sites can be created or destroyed, but only in the stoichiometric ratio of the host!

An Example Point Defect Reaction

OOx ⇐⇒ 1/2O2(g) + VO

•• + 2e'

Kred pO2-1/2 = [VO

••] n2

n = 2 [VO••] or [VO

••] = n/2

n = (2Kred )1/3 pO2-1/6

ZnO

Point defects in n-type TCOs

log pO2 log [D]

n=2[VO••] n=2[VO

••] n=[D•] n=[D•]

-1/6

-1/2 -2

OOx ↔ 1/2O2(g) + VO

•• + 2e'

intrinsic extrinsic intrinsic extrinsic

Is this the whole story?

1 24

0.1

1

10

-16 -12 -8 -4 0

Po2 (atm)

Nor

mal

ized

con

duct

ivity

(σ/σ m

ax)

Norm

alized thermopow

er (Q/Q

max )

10 1010 10

81−

10

241

121

41−

0.1

1

10

-6 -4 -2 0

Po2 (atm)

Nor

mal

ized

con

duct

ivity

(σ/σ m

ax)

Norm

alized thermopow

er (Q/Q

max )

10 1010 10

81−

241

41−

121

1 12

1 12

Bulk ITO at 800oC

Nano ITO at 500oC

Q ∝ m*/n2/3; m* ∝ n1/3

Q ∝ n-1/3

Electrical Properties

1 24

24

Formation of point defect associates:

Reactiion: (2Sn•In

Oi'')x → ½ O2 (g) + 2Sn•In + 2 e'

n ∝ pO2 -1/8 Kröger-Vink Sn.

In represents a tin (on an indium site) with 1+ charge

notation Oi” represents an oxygen interstitial with 2– charge

OO represents an oxygen (on an oxygen site)

( )x represents a neutral defect

1 G Frank and H Köstlin, Appl Phys A, 27 (1982), 197

The Frank-Köstlin Cluster

ITO Defect Chemistry

• Rietveld analysis of combined neutron and X-ray diffraction patterns (bulk and nano ITO) Sn-to-oxygen interstitial ratio ~2:1

• EXAFS studies of bulk and nano ITO specimens – Sn coordination number > 6 – First near neighbor distances of Sn similar to that

in SnO2

Evidence for Defect Associates in ITO

00 )( IIxeII t ∝⇒= − µµ

Determines: Oxidation/charge state Local structure

Atomic distances Coordination numbers

Element specific Sensitive at low concentrations

X-ray Absorption Fine Structure

Defect Chemistry of n-type TCOs (Cont’d)

1175°C

CdO (rocksalt) o-Cd2SnO4

(Sr2PbO4) CdSnO3

(o-Perovskite)

CdIn2O4 In2-2x(Cd,Sn)2xO3 (0 < x < 0.34)

(bixbyite)

ITO

SnO2 (rutile)

(1-x)CdIn2O4 – (x)Cd2SnO4 (0 < x < 0.75)

(spinel)

InO1.5 (bixbyite)

rutile + bixbyite

spinel + rocksalt

spinel + bixbyite

D. R. Kammler, T. O. Maosn, K. R. Poeppelmeieir, J. Am. Ceram. Soc., 84, 1004 (2001)

The Cd-In-Sn System

Compound Structure Solubility

CdSnO3 o-Perovskite 4.5%

In2O3 Bixbyite 34%

CdIn2O4 Spinel 75%

Isovalent substitutions: [D•] = [A′]

In2O3 + CdCdx + SnSn

x InCd• + SnIn ′ + CdSnO3

CdO + SnO2 +2InInx CdIn ′ + SnIn

• + In2O3

Size-matched substitutions:

A0.82

2A0.69A0.95

2rr

A0.80r42

3SnCd

In =+

=+

≈=++

+

Extended Solubility Ranges

[In] = 0.66

InO1.5

CdO SnO2

Tie lines connecting bixbyite & spinel phase fields

Tie lines connecting bixbyite & rutile

phase fields

Line of perfect stoichiometry In2-2xCdxSnxO3

[SnIn•] - [CdIn′] = n

Spinel

Biphasic Studies

0

1

2

0 0.1 0.2 0.3 0.40

1

2

0 0.1 0.2 0.3 0.4

log(σ S

n ric

h/σC

d ric

h)

x in In2-2x(Sn,Cd)2xO4 lo

g(σ r

educ

ed/σ

as-fi

red)

x in In2-2x(Sn,Cd)2xO4

phase field width sensitivity to reduction

Conductivity Variations

[In] = 0.66

InO1.5

CdO SnO2

Tie lines connecting bixbyite & spinel phase fields

Tie lines connecting bixbyite & rutile

phase fields

Line of perfect stoichiometry In2-2xCdxSnxO3

[SnIn•] - [CdIn′] = n

Spinel

Self-Doped Materials

self-doped TCOs

-600 1 2

Free

Ene

rgy

D/A < 1 D/A > 1

lightly co-dopedbixbyite

heavily co-dopedbixbyite

-600 1 2

Free

Ene

rgy

D/A < 1 D/A > 1

lightly co-dopedbixbyite

heavily co-dopedbixbyite

Inherent Off-Stoichiometry

In2O3(ZnO)k

T = 1275 ˚C

Zn2SnO4

The Zn-In-Sn-O System

oxygen sensitive regime n= f (Oi)= f (PO2)

exhaustion regime

(2SnIn• Oi’’)x 2SnIn

• +2e’ +1/2O2

In1.2Sn0.40Zn0.40O3

or ZITO40 co-doped In2O3

Bixbyite structure

The -1/8 slope attributable to (2SnIn

• Oi’’)x complex

In1.2Sn0.40Zn0.40O3

or ZITO40

750 °C

Log PO2

Log

σIT

O

-1/8

(2SnIn• Oi’’)x 2SnIn

• +2e’ +1/2O2

In Situ Conductivity: In1.2Zn0.4Sn0.4O3

1175°C

CdO (rocksalt) o-Cd2SnO4

(Sr2PbO4) CdSnO3

(o-Perovskite)

CdIn2O4 In2-2x(Cd,Sn)2xO3 (0 < x < 0.34)

(bixbyite)

ITO

SnO2 (rutile)

(1-x)CdIn2O4 – (x)Cd2SnO4 (0 < x < 0.75)

(spinel)

InO1.5 (bixbyite)

rutile + bixbyite

spinel + rocksalt

spinel + bixbyite

D. R. Kammler, T. O. Maosn, K. R. Poeppelmeieir, J. Am. Ceram. Soc., 84, 1004 (2001)

The Cd-In-Sn System

Evidence for D/A Excess in Spinels

Cd2SnO4 Theory Predictions • n-type character not attributable to typical

intrinsic donors: – VO has a small ΔHf, but lies deep in the gap – Cdi is a shallow donor, but ΔHf is large

• Antisite defects are shallow donors (SnCd•)

and acceptors (CdSn′), but the former has a lower ΔHf

• The net result: [SnCd•] > [CdSn′]; Cd2SnO4 is

consistently n-type

“Self-doping of cadmium stannate in the inverse spinel structure," S. B. Zhang and S. –H. Wei, Appl. Phys. Lett. 80, 1376 (2011).

What About p-Type TCOs? Stay tuned…

Conclusions

• With the exception of ZnO, the best TCOs are high octahedral site density crystal structures

• They are n-type only • The “basis” cations are In, Sn, Zn, Cd and Ga • Their defect chemistries can be complicated

(not just oxygen vacancies) • Phase diagrams play an important role • If you can make a comparable p-type TCO…